ADJUSTMENT OF OBSERVATIONS. 



69 



a stream with the critical bottom velocity does 

 not transport debris and therefore does not 

 establish its own slope of bed. The slopes 

 artificially prepared were of necessity imper- 

 fectly adjusted. In the next place, the pre- 

 pared slopes did not imitate the natural 

 diversity of detail but were plane. When a 

 current of competent velocity passed over one 

 of them it immediately began to shape the bed 

 into dunes, and as the modeling proceeded the 

 activity of transportation increased. After the 

 dunes were formed, a smaller general velocity, 

 or a less discharge, or a lower general slope 

 was competent. The experiments being made 



in sets, the first of a set gave a result from the 

 plane bed and the others from a more or less 

 diversified bed. 



In the third place, the particles composing 

 one of the experimental grades of debris were 

 not of uniform mobility. Not only were they 

 of diverse size, as indicated by the "range of 

 fineness," but they were different in shape and 

 in specific gravity, so that some were able to 

 resist a considerably stronger current than 

 others. The competent slope for the least 

 mobile particles was materially steeper than 

 that for the most mobile, and no mode of 

 gaging average mobility was discovered. 



TABLE 9. Experimental data on competent slope. 



When an experiment was begun with a velocity 

 well below competence, and the velocity was 

 gradually increased, the first movement de- 

 tected would be the saltation of some small or 

 light particle, and then the number of particles 

 moving would gradually grow with the quicken- 

 ing of current. 



An attempt to correlate the " notes on move- 

 ment of de'bris" in Table 10 for a discharge of 

 0.363 ft. 3 /sec. gave the following values of 

 water slope for equivalent phases of movement: 



Grade (B) (C) (D) (E) (G) 



Slope of water (per 

 cent) 0.03 0.03 0.06 0.10 0.93 



By assuming the power -function S c = a F 2 ", or 

 log S c = log a + n log F 2 , and plotting log S c in 



relation to log F 2 , the value found for n is about 

 0.5, but it has a large uncertainty. The 



oo 



S-i 



01234 

 LogF 



FIGURE 25. Logarithmic plot of competent slope in relation to fineness 

 ofdeVis. 



plot, figure 25, shows the competent slope for 

 grade (E) smaller than would be indicated by 



